☰

Cavidi AB

From research-grade chemistry to global diagnostic kits

Industry
In-vitro Diagnostics
Headquarters
Uppsala, Sweden
Public information as of
February 2026

A4BEE prepared this analysis from publicly available sources. It reflects our own reading of Cavidi AB's published strategy and is not endorsed by, or produced in cooperation with, Cavidi AB. Company website

Strategic priorities

Cavidi AB was founded in Uppsala in 1984 as a University spinoff and built its business around Binding Oligo Ladder Detection (BOLD), a signal-amplification chemistry that delivers up to 200x amplification on standard ELISA readers rather than on the proprietary $250,000 platforms used by competitors such as Quanterix. In January 2025 the company raised 7.8 million euros of debt and equity to commercialize its Exazym reagent kits globally, under a new CEO, Jonathan Royce, appointed in September 2025 with industrial track records at GE Healthcare, Millipore and Bio-Works.

The immediate operational shift is from laboratory-scale Research Use Only (RUO) kit assembly to industrial production of In-Vitro Diagnostic (IVD) kits under the EU In-Vitro Diagnostic Regulation (IVDR 2017/746). Reagent formulation involves reverse-transcriptase polymerases, oligo-dT primers and specific DNA:RNA templates whose concentrations have to be held within tight bounds across every well of a 96-well plate. Manual pipetting and concentration management, acceptable for a research product, set the company's tolerance for batch variance and documentation discipline.

The product itself is platform-agnostic: BOLD works on Luminex, Gyrolab and standard ELISA readers, and that openness is Cavidi's commercial position against proprietary lock-in. The same openness means assay data lands in different vendor formats, and bringing those formats under one model is the precondition for the unified dashboards the company's CEO has said he wants to build for clients of the Bespoke Service. Distribution runs through named partners in China (Dakewe), India (Invenio Life) and North America, so every technical decision has to hold across geographies.

Three capabilities sit underneath the stated scale-up: closed-loop reagent formulation that holds attomole-level precision on the production line, an industrial data layer that pulls assay results from heterogeneous readers into one model, and an IVDR-aligned documentation and quality system that produces evidence continuously rather than assembling it per audit. None of these is purely a software decision — together they decide whether Exazym reaches global clinical use as a registered IVD kit or remains an RUO product with limited distribution.

Challenges we see

  • Operations Manufacturing

    Holding attomole-level precision on the production line

    Exazym reagent kits combine reverse-transcriptase polymerases, oligo-dT primers and specific DNA:RNA templates in a 96-well format whose 200x signal amplification depends on precise reagent concentrations and incubation times.

    Where reagent concentrations are set by hand at attomole sensitivity, the practical unit of review becomes the individual well rather than the batch, and the production line becomes the place where precision has to be generated rather than confirmed afterwards.

  • Digital Integration

    Unifying data from heterogeneous immunoassay readers

    BOLD works with Luminex, Gyrolab and standard ELISA readers, and the company distributes through named partners in China, India and North America. Assay data lands in vendor-specific output formats rather than in a shared internal model.

    Where assay results stay inside vendor-specific reader software, the practical integration point is the data model rather than the instrument, and an agreed ontology for assay, specimen, result and reader becomes the precondition for any fleet-level view.

  • Compliance Regulatory

    Moving from RUO documentation to IVDR evidence

    Cavidi is transitioning its Exazym kit line from Research Use Only to In-Vitro Diagnostics under EU IVDR 2017/746, which raises the bar for audit trails, data integrity, change control and performance evaluation compared with RUO.

    Where evidence was compiled for periodic review under RUO, the IVDR expectation is continuous generation of audit-ready records, so the documentation system becomes part of the production line rather than a step that follows it.

  • Workforce Operations

    Specialised bioprocess talent in a competitive cluster

    Uppsala's biotech cluster competes for specialised bioprocess engineers against larger Contract Development and Manufacturing Organisations (CDMOs) and established industry players. The 7.8 million euro 2025 funding round commits the company to a scale-up phase that depends on a small core of skilled operators.

    Where the same team is expected to deliver more kits on a tighter regulatory clock, the practical surface is the operator interface rather than the underlying chemistry, and procedures that depend on a single expert's recall become procedures that have to be encoded in the system itself.

  • Digital Operations

    Guiding long incubation steps without tying scientists to the bench

    Exazym Instructions for Use describe multi-step manual work including reagent preparation, template mixing and a 10 to 30 minute incubation phase that currently requires continuous bench monitoring.

    Where incubation steps are supervised by an operator standing at the bench, the practical question becomes how the same reaction is supervised remotely with the same confidence, and the answer depends on the time-series data the system produces during the run rather than on the scientist's presence.

Opportunities, by urgency and business impact

Each bubble is one opportunity, numbered to match the list below. Further right means it bites sooner; higher means a bigger effect on the business. A bigger bubble means a bigger implementation effort.

Source: A4BEE analysis of public sources

  1. Holding batch-to-batch consistency at attomole sensitivity

    Exazym reagent formulation involves reverse-transcriptase polymerases, oligo-dT primers and specific DNA:RNA templates at concentrations that determine the 200x signal amplification that defines BOLD's commercial value. Manual pipetting and concentration management keep the variance per well at levels acceptable for an RUO product but not for an IVD product distributed globally.

    Closed-loop dosing and IoT-connected thermal control during reagent formulation and incubation lets concentrations and temperatures be measured and adjusted at the well rather than at the batch, so the consistency the chemistry needs is generated by the line rather than confirmed afterwards.

    • Cavidi, Exazym Polymerase Reaction Kit Instructions for Use, December 2023
    • Jonathan Royce CEO appointment press release, Cavidi, September 2025

  2. Bringing Luminex, Gyrolab and ELISA readers under one data model

    BOLD assay results land in vendor-specific output formats across Luminex, Gyrolab and standard ELISA readers, and there is no shared internal model that lets results from different platforms be compared, monitored or aggregated. The same gap affects partner laboratories that send Exazym data back to Cavidi for the Bespoke Service.

    An ontology-based data platform that defines assay, specimen, result, instrument and lot once and ingests vendor outputs through OPC UA (Open Platform Communications Unified Architecture) or file-based adapters lets internal teams and partner labs query one model instead of reconciling per report.

    • Cavidi homepage, Exazym product description
    • Cavidi_DeepResearch, A4BEE Strategic Account Intelligence, February 2026

  3. Producing IVDR audit-ready records as part of the run

    The transition from Research Use Only to In-Vitro Diagnostics under EU IVDR 2017/746 raises documentation requirements for data integrity, audit trails, change control and performance evaluation beyond what manual logs and Excel-based QC tracking can support. IVDR audit findings on a growing IVD kit line would be commercially costly.

    Capturing reagent preparation, run conditions and QC results as electronic records with their own audit trail, anchored to the batch record from the start, produces IVDR-grade evidence as part of the run and shortens the path between production and inspection.

    • Cavidi_DeepResearch, A4BEE Strategic Account Intelligence, IVDR transition section, February 2026
    • Cavidi about page, Quality & Projects function, February 2026

  4. Speeding up BOLD transfer to new partner analytes

    The Bespoke Service extends BOLD into new analytes on partner immunoassay platforms through manual integration and optimisation that takes weeks to months per analyte, keeping the service high-margin but low-throughput.

    A computational model of BOLD polymerization kinetics lets assay performance for a new analyte be predicted before wet-lab work begins, so partner projects move from a calendar-bound optimisation cycle to a search over predicted configurations that the wet-lab team then validates.

    • Cavidi homepage, Exazym product description
    • Cavidi Bespoke Service description, Cavidi press materials, 2025

  5. Monitoring laboratory conditions for attomole-level work

    BOLD's attomole-level sensitivity makes assay results vulnerable to temperature drift, humidity change and airborne particulates in the production environment, and the current review of those conditions depends on manual checks rather than continuous measurement.

    A continuous environmental sensing layer over the laboratory and production spaces provides the data needed to qualify the production environment for IVDR and to detect drift before it affects the next batch, rather than after.

    • Cavidi_DeepResearch, A4BEE Strategic Account Intelligence, environmental control section, February 2026

What we'd propose

  • Digital CDMO

    Closed-loop kit production line with continuous process data

    We retrofit the Exazym kit production line with IoT-connected dosing, thermal control and vision-based verification, and we stream process values into a per-batch record, so that reagent concentrations and incubation conditions are set by the line rather than by hand.

    • IoT-connected dosing and thermal control

      Concentrations and temperatures on the line

      Connect precision dosing pumps and thermal controllers through OPC UA (Open Platform Communications Unified Architecture) or MQTT so reagent volumes and incubation temperatures are measured at the well rather than estimated, with feedback loops adjusting setpoints as the run progresses.

    • Vision-based run verification

      A camera on every well

      Deploy compact vision stations that capture reagent addition and fill-level anomalies during formulation, flagging deviations to the operator interface while the batch is still in production rather than at end-of-line QC.

    • Per-batch process record with lineage

      Evidence assembled as you go

      Stream monitored parameters into a per-batch record with lineage back to the sensor or camera that produced each value, so the evidence behind a release decision is generated by the line and stays inspectable end to end.

    • Reagent precision is generated by the line at the well, not confirmed afterwards at the batch.
    • The same process data serves manufacturing, quality and regulatory, so each release is assembled from the same record.
    • Changeover between kit variants is faster because the production system carries the recipe and the verification step in one place.

    We've done this before

    See how we do Digital CDMO
  • Enterprise AI

    Ontology-based data platform for multi-reader diagnostics

    We build an ontology-based data platform that defines assay, specimen, result, instrument and lot once, ingests outputs from Luminex, Gyrolab and ELISA readers through vendor-neutral adapters, and exposes the combined data set to internal teams and Bespoke Service partners.

    • Shared diagnostics ontology

      One agreed set of terms

      Define assay, specimen, result, instrument, lot and run as explicit entities with agreed relationships, so a query written once returns comparable answers across reader types and across partner laboratories instead of two dialects of the same table.

    • Reader and partner data ingestion

      Loading every reader format

      Build ingestion adapters for Luminex, Gyrolab and standard ELISA readers, and an API gateway that lets Bespoke Service partners push results back into the same model with schema validation at the boundary so bad records fail loudly.

    • Dashboards and partner-facing retrieval

      Questions answered without new extracts

      Expose the model through internal dashboards and a partner-facing retrieval layer, so commercial, scientific and operations teams can ask questions of the combined data set without commissioning a new extract for each one.

    • Integration work is done once against the shared model instead of once per point-to-point interface.
    • Bespoke Service partners attach their data to a known model rather than triggering a new migration for every analyte.
    • New assay types and new readers attach to the ontology instead of becoming a separate data silo.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Paperless, IVDR-aligned laboratory and production records

    We replace paper Instructions for Use, manual logs and Excel-based QC tracking with a tablet-guided digital workflow that captures every step of an Exazym run as electronic record, with timestamps, change tracking and audit trail ready for IVDR inspection.

    • Tablet-guided protocol execution

      Step-by-step on the device

      Replace the static Exazym Instructions for Use with an interactive tablet application that guides scientists through reagent preparation, template mixing and incubation, recording execution data and time stamps as the run progresses.

    • Automatic audit trail

      Records that hold up to inspection

      Generate tamper-evident electronic records with versioning and electronic signatures that meet IVDR 2017/746 expectations for data integrity, change tracking and regulatory submission, replacing the manual logs and Excel-based QC that currently hold the documentation load.

    • Remote incubation supervision

      Watching runs from anywhere

      Surface the run state of every active 10 to 30 minute incubation on a single mobile or web dashboard, so a scientist can supervise several runs in parallel without being at each bench for the full reaction window.

    • Documentation effort moves from compiling for inspection to generating during the run.
    • A single record set serves the laboratory, manufacturing and the IVDR file, so the same evidence is reviewed once.
    • Run supervision stops being a single-operator commitment at a single bench.

    We've done this before

    See how we do Digital Lab
  • Digital Lab

    Computational model of BOLD polymerization kinetics

    We build a computational model of BOLD's reverse-transcriptase polymerization step that predicts assay performance for new partner analytes before wet-lab work, so the Bespoke Service moves from calendar-bound optimisation to a search over predicted configurations.

    • Kinetic model of the polymerization step

      The chemistry in software

      Model reverse-transcriptase activity, oligo-dT binding and signal amplification kinetics as a parameterised simulation that runs against a new analyte's expected binding profile, with documented assumptions and known operating ranges.

    • Partner-facing assay configurator

      Predicted performance up front

      Expose the kinetic model through a configurator that lets Bespoke Service clients enter new analyte parameters and receive predicted amplification curves and operating windows before committing to wet-lab development.

    • In-silico protocol validation

      Validate before the wet-lab starts

      Run simulated protocol validation to identify the operating range that the wet-lab team should target, so that IQ/OQ/PQ (Installation, Operational and Performance Qualification) work starts from a known-good window rather than from a guess.

    • Bespoke Service timelines shorten because the chemistry is explored in software before wet-lab work begins.
    • Partner projects arrive at the wet-lab with a predicted operating window rather than a blank sheet.
    • Validation work focuses on confirming a known range rather than on discovering one.

    We've done this before

    See how we do Digital Lab
  • Agents

    AI agents for IVDR documentation work

    Narrow, reviewable agents that take the repetitive part of IVDR documentation work: drafting technical files and performance evaluation summaries from source records, checking a document against its template before review, and finding every controlled document that a standards change touches. A named person approves every output.

    • Drafting from source records

      First drafts from system data

      Generate the first draft of IVDR technical file sections, performance evaluation summaries and change-control documents directly from the underlying laboratory, production and QC records, so the author edits and judges rather than assembles text from scratch.

    • Template and completeness checking

      Gaps found before review

      Check a submitted document against the relevant IVDR template and the site's own checklist, returning missing or inconsistent sections before the document enters the human review queue.

    • Change impact search across the document set

      Which documents a change touches

      When a standard, method or specification changes, retrieve every controlled document that references it and rank them by how directly they are affected, so the update scope is known on day one.

    • Technical file review queues move faster because documents arrive complete and supported by links to source records.
    • The scope of a standards or method change is established by search rather than by recollection.
    • Every agent output is traceable to the source records it came from and signed off by a named reviewer.

    We've done this before

    See how we do Agents

Digital maturity: today and target

Scored out of 100 across six dimensions. The target is what Cavidi AB's own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Manufacturing line process data 25 → 75
Exazym kit assembly is currently manual or semi-automated. The transition to a closed-loop production line with streamed process data is the precondition for IVDR-grade batch consistency and is ahead of the platform rather than behind it.
Data integration 20 → 80
BOLD produces data on Luminex, Gyrolab and standard ELISA readers in vendor-specific formats, with no shared ontology across platforms or partner laboratories. The ontology and ingestion work that brings the data into one model is still to be built.
Process digitalization 30 → 85
Paper Instructions for Use, manual logs and Excel-based QC tracking currently hold the laboratory and production documentation load. A tablet-guided digital workflow that captures execution data in real time is the gap between current and IVDR-ready operations.
Quality and compliance 35 → 90
RUO-era documentation is in place but does not meet IVDR 2017/746 expectations for audit trail, data integrity and performance evaluation. The compliance system has to move from assembling evidence per audit to generating it continuously as part of the run.
Predictive analytics 15 → 70
There is no predictive model of BOLD polymerization kinetics or of laboratory environmental drift today. Both are stated needs: partner assay transfer would benefit from kinetic prediction, and IVDR-grade environmental qualification depends on continuous measurement rather than spot checks.
IT/OT convergence 20 → 75
Reader outputs, dosing pumps, thermal controllers and laboratory sensors are not connected to enterprise systems today, so production and quality decisions are made on partial information. The convergence work covers both the line-side connection and the cloud-side retrieval.

Check this yourself

Our Service Portal has free self-assessments and market comparisons. These are the ones that line up with what we've read above — no sales call required.

Think we've read this right?

Talk to us

Related reading

This is an independent analysis prepared by A4BEE from publicly available information as of February 2026. It reflects A4BEE's own interpretation and opinion, is not affiliated with, endorsed by, or verified with Cavidi AB, and may be incomplete or inaccurate. All company names and trademarks are the property of their respective owners. To request a correction or removal, contact [email protected].